A general apparatus for recognizing a driver's voice, such as a handsfree apparatus or the like of a vehicle is implemented by using a single microphone having directivity toward a driver's seat direction or by using an array microphone for beam-forming toward the driver's seat direction in order to minimize an influence of noise included in the driver's voice.
However, since the driver's voice is generally obtained in a state in which a sitting height of the driver, a position of the driver, and the like are not considered, it may be insufficient to provide high quality call environment.